AI 无需见过世界也能学会预测世界

编辑:Lisa Lock,审校:Robert Egan
本文已按照 Science X 的编辑流程与政策进行审校。编辑在确保内容可信度的同时,重点标注了以下属性:
添加到首选来源
零数据的自我对弈预训练。图片来源:arXiv (2026)。DOI: 10.48550/arxiv.2609.30063
我们所知的几乎每一个人工智能系统都以同样的方式起步:让它接触现实世界中的海量信息——书籍、网站、图像和录音——并训练它识别其中的模式。一项由特拉维夫大学和斯坦福大学研究人员参与的新研究提出了另一种可能:即使系统训练所用的全部数据都完全由系统自身从零生成,它是否仍能发展出通用学习能力?
研究人员证明这是可行的,这标志着朝着长期以来被视为 AI 训练圣杯的目标迈出了重要一步。该研究在特拉维夫大学 Blavatnik 计算机科学学院的 Yoav Levine 博士指导下完成,是 Kfir Dolev 博士后研究的一部分,并与斯坦福大学的研究人员合作开展。该研究已发布在 arXiv 预印本服务器上。
两个 AI 模型共同学习
研究人员提出了一种训练方法,让两个模型从零开始、共同学习。第一个模型生成能产出符号序列的简短计算机程序;第二个模型则被训练来预测每个序列中的下一个符号。生成练习题的模型会收到反馈,了解这些题目对学习有多大帮助,并逐渐更擅长挑选与学习模型当前发展中的能力相匹配的挑战。
决定性的测试出现在训练之后:模型从这些人工序列中学到的东西,是否也能帮助它预测训练中从未接触过的真实世界数据?研究人员评估了它在文本、图像、声音、音乐、代码和 DNA 序列上的表现。
他们发现,随着投入训练过程的算力资源增加,模型在多个领域的预测能力持续提升。该模型还能在没有任何额外训练的情况下,从测试时提供给它的示例中推断出规则。
当 AI 自己创造课程时,会发生什么?
其中最引人注目的发现之一,涉及该系统为自己生成的学习材料。在训练过程中,这些材料里出现了人们熟悉的数学模式,包括斐波那契数列。
换句话说,该系统学会了寻找生成结构化序列的方法,使第二个模型能够从中受益。在另一项实验中,研究人员发现,用这种新方法进行预训练,能让模型在随后接触真实世界数据时学得更快。
这对 AI 的未来意味着什么?
这项研究为人工智能研究中的一些根本问题提供了初步答案:智能是否存在不依赖任何外部数据的“通用”成分?模型的部分学习能力是否源自无需事先接触真实世界信息就能获得的通用模式?研究结果表明,这确实是可能的。不过,研究人员强调,该方法无法在不接触现实的情况下让系统学会关于现实的事实。更确切地说,它或许最终能帮助 AI 系统更高效地从真实世界数据中学习。
Levine 总结道:“多年来,普遍的看法是,AI 模型要想学到任何有用的东西,首先必须被喂入海量的真实世界数据。在这项研究中,我们推翻了这一根本假设,并证明即使模型仅在学习过程本身生成的数据上训练,也能获得通用的预测能力。
“这仍然是一个涉及小模型的可行性验证,但这些发现为未来研究提出了引人入胜的问题:当我们用这些方法强化 AI 时,会涌现出哪些普适的数学规律?在不接触外部世界信息的情况下,我们能把 AI 推进到多远?我们能否让模型通过自身生成的经历,为更有效地从现实中学习做好准备?”
Aditya Cowsik 等,《Self-Play Pretraining with Zero Data》,arXiv(2026)。DOI: 10.48550/arxiv.2609.30063
期刊信息:arXiv
这篇报道背后是谁?
艺术史学士,物质文化硕士。曾任博物馆编辑、急救员和移植协调员。自 2021 年起为 Science X 担任编辑。完整简介 →
数学生物学学士,创意写作硕士。游历广泛,对科学与语言有独到见解。完整简介 →
引用:AI 学会在从未见过世界的情况下预测世界(2026 年 11 月 5 日),2026 年 11 月 6 日检索自 https://techxplore.com/news/2026-11-ai-world.html
本文件受版权保护。除为私人学习或研究目的的合理使用外,未经书面许可,不得复制本文件的任何部分。内容仅供参考。
获取即时摘要文本(Gist)
在内部生成的符号序列上进行训练,使小型 AI 模型能够跨多个领域预测未见过的真实世界数据。随着训练算力的增加,性能有所提升,而这种预训练加速了后续从真实数据中的学习。这些结果支持的是可迁移的预测能力,而非无需接触就获得真实世界事实。
本摘要由 LLM 自动生成。完整免责声明
译文由英文原文机器翻译,术语按行业惯例统一,可在文末展开原文核对。
查看英文原文(AI learns to predict the world without ever seeing it)
AI learns to predict the world without ever seeing it
AI learns to predict the world without ever seeing it
edited by Lisa Lock, reviewed by Robert Egan
This article has been reviewed according to Science X's editorial process and policies. Editors have highlighted the following attributes while ensuring the content's credibility:
Add to Preferred Sources
Self-play pretraining with zero data. Credit: arXiv (2026). DOI: 10.48550/arxiv.2609.30063
Almost every artificial intelligence system we know begins in the same way: it is exposed to vast amounts of information from the real world—books, websites, images and recordings—and trained to identify patterns within it. A new study involving researchers from Tel Aviv University and Stanford University raises another possibility: Can a system develop general learning capabilities even when all the data it trains on is generated entirely from scratch within the system itself?
The researchers have shown that it can, marking a significant step toward a goal long considered the holy grail of AI training. The study was conducted under the supervision of Dr. Yoav Levine of the Blavatnik School of Computer Science at Tel Aviv University, as part of the postdoctoral research of Dr. Kfir Dolev, in collaboration with researchers from Stanford University. The study is available on the arXiv preprint server.
Two AI models learning together
The researchers presented a training method in which two models start from scratch and learn together. The first generates short computer programs that produce sequences of symbols. The second is trained to predict the next symbol in each sequence. The model generating the exercises receives feedback on how much they contribute to learning and gradually becomes better at selecting challenges suited to the learning model's developing capabilities.
The decisive test came after training: Would what the model had learned from these artificial sequences also help it predict real-world data it had never encountered during training? The researchers evaluated its performance on text, images, sounds, music, code and DNA sequences.
They found that as more computational resources were devoted to the training process, the model's predictive capability improved consistently across many domains. The model was also able to infer rules from examples presented to it during testing, without any additional training.
What happens when AI creates its own lessons?
One of the most intriguing findings concerned the learning material the system generated for itself. During training, familiar mathematical patterns appeared in this material, including the Fibonacci sequence.
In other words, the system learned to find ways to generate structured sequences from which the second model could benefit. In another experiment, the researchers found that pretraining with the new method helped the model learn faster when it was subsequently exposed to real-world data.
What could this mean for the future of AI?
The study offers an initial answer to fundamental questions in artificial intelligence research: Does intelligence have a "universal" component that does not depend on any external data? Does part of a model's learning capacity stem from general patterns that can be acquired without prior exposure to real-world information? The findings indicate that this is indeed possible. However, the researchers emphasize that the method cannot teach a system facts about reality without exposing it to that reality. Rather, it may eventually help AI systems learn from real-world data more efficiently.
Levine concludes, "For years, the prevailing assumption was that for an AI model to learn anything useful, it first had to be fed enormous amounts of real-world data. In this study, we disproved this fundamental assumption and showed that a model can acquire general predictive skills even when it is trained only on data generated within the learning process itself.
"This is still a proof of concept involving small models, but the findings raise fascinating questions for future research: What universal mathematical patterns will emerge as we strengthen AI using these methods? How far can we take AI without exposing it to information from the outside world? And can we prepare models to learn more effectively from reality through experiences they have generated for themselves?"
Aditya Cowsik et al., Self-Play Pretraining with Zero Data, arXiv (2026). DOI: 10.48550/arxiv.2609.30063
Journal information:arXiv
Who's behind this story?
BA art history, MA material culture. Former museum editor, paramedic, and transplant coordinator. Editing for Science X since 2021. Full profile →
Bachelor's in mathematical biology, Master's in creative writing. Well-traveled with unique perspectives on science and language. Full profile →
Citation: AI learns to predict the world without ever seeing it (2026, November 5) retrieved 6 November 2026 from https://techxplore.com/news/2026-11-ai-world.html
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.
Get Instant Summarized Text (Gist)
Training on internally generated symbol sequences enabled small AI models to predict unseen real-world data across multiple domains. Performance improved with more training compute, and this pretraining accelerated subsequent learning from real data. The results support transferable predictive skills, not acquisition of real-world facts without exposure.
This summary was automatically generated using LLM. Full disclaimer
This site uses cookies to assist with navigation, analyse your use of our services, collect data for ads personalisation and provide content from third parties. By using our site, you acknowledge that you have read and understand our Privacy Policy and Terms of Use.
原文为英文,正文由自动抓取并保留原文。
原文发布于 TechXplore(AI learns to predict the world without ever seeing it),标题、摘要与正文由机器翻译,术语按行业惯例统一,未经人工逐句校订。
